On Tue, 2010-03-09 at 19:57 +0100, Attila Csipa wrote: > On Tuesday 09 March 2010 15:54:54 Ville M. Vainio wrote: > > On Tue, Mar 9, 2010 at 2:07 PM, Giovanni Bajo <[email protected]> wrote: > > > It might make more sense to have a way to export PyQt classes to QML, > > > that is implement QMLElement in Python rather than in C++, and make them > > > available to QML itself. > > > > This is pretty much what I'm thinking of. Python would take the part > > of C++ here, not Javascript. > > If I understand correctly what you want to do, it has little to do with PyQt, > it would be more of a PyQML custom binding... Which is interesting, but > sounds > like quite a niche to me - it will be hard to avoid Qt parts, and at that > point it's much easier to go via the 'main' PyQt path. But, I'm no QML expert > (yet :), so take this with a grain of salt, I don't intend to discourage > anyone. What might be related, but I also don't know enough about, is what > the > current state of QtScript is. Is currently it possible to operate on Python > objects from QtScript ? 'Cause if it is, QML should not be far off.
Your two examples are unrelated. QtScript creates an automatic binding between Javascript and C++ using QObject introspection features. Thus, if you create a Python QObject, explicitly define signals/slots/properties at the QMetaObject level through PyQt decorators, the resulting object will be accessible from QtScript, with QtScript totally ignoring it is a Python object under the hood. For a C++ object to be exported to QML, on the other hand, you need an additional set of meta-definitions which are currently unavailable from Python. I'm not even sure if those meta-definitions resolves to *runtime* data that can be dynamically build, as it would be required by PyQt. Surely, there is more work to be done on the PyQt side. -- Giovanni Bajo Develer S.r.l. http://www.develer.com _______________________________________________ PyQt mailing list [email protected] http://www.riverbankcomputing.com/mailman/listinfo/pyqt
